5 Meta Ads mistakes I see every week

Most Meta ad accounts I look at don’t have a creative problem or a budget problem. They have a setup problem. The same five mistakes come up again and again, and each one quietly tells Meta to spend your money on the wrong thing.

01 · Boosting posts instead of running campaigns

The Boost button is built around a post: it puts that post in front of more people. That’s fine for awareness. It isn’t built for enquiries.

Meta delivers ads to the people most likely to take the action you optimise for. If the goal is engagement, you get likes and comments, often from people who will never buy. For enquiries, create a campaign in Ads Manager with a Leads goal (an instant form, a website conversion or a message), so delivery is optimised for the thing that actually pays you.

Fix: move the budget from boosts into one Ads Manager campaign with a lead or message goal.

02 · No Pixel, or a Pixel that only sees page views

If Meta can’t see a lead happen, it can’t find more people like the ones who became leads. A Pixel that only fires PageView is nearly as bad as none.

Set up the events that matter, usually Lead and Contact, and send them from both the browser (Pixel) and your server (the Conversions API), so fewer are lost to ad blockers and browser privacy settings. When both send the same event, Meta removes the duplicate by matching the event_name and a shared event_id, so give every event a unique ID (Meta’s deduplication guide).

Fix: open Events Manager, test each event, and check that browser and server events show as deduplicated.

03 · One ad, judged after two days

One ad is not a test, and two days is rarely enough to judge it. Meta’s own guidance is to wait for roughly 50 optimisation events after the last significant edit before reading costs, and a significant edit includes pausing the ad set or changing its optimisation event, audience or creative (Meta Business Help Center). Every time you tweak, the clock restarts.

Fix: launch three to six ads with different hooks against one goal, then leave them alone long enough to collect results. Change one thing at a time.

04 · Lead forms built for volume

Instant forms are easy to submit, which is exactly why they can fill up with people who don’t remember filling them in. Two settings change lead quality more than anything else: the form type and a qualifying question.

A higher-intent form adds a review step before submitting. A custom question, like “When do you want to start?” or “Which area are you in?”, filters out people who were only curious and tells you who to call first.

Fix: switch to a higher-intent form and add one question that a serious customer can answer easily.

05 · Replying the next day

The cheapest lead in the account is wasted if nobody replies until tomorrow. By then the person has usually messaged someone else. This is the mistake that no amount of ad optimisation can fix.

If leads arrive on WhatsApp, the timing also matters for cost: you can reply freely within 24 hours of the customer’s last message, and leads from click-to-WhatsApp ads can unlock a longer free window when you reply inside the first 24 hours (WhatsApp Business Platform pricing).

Fix: decide who answers leads and how fast, before the campaign starts. Send form leads straight to a phone, not just an inbox.

The order to fix them in

Start with tracking (02), because nothing else can be measured without it. Then the campaign goal (01) and the form (04), then creative testing (03). Reply speed (05) costs nothing to fix and often matters most.

If you want a second pair of eyes on your own account, the free 15-minute review covers exactly these five things. Or read how I run Meta Ads.

Meta optimises delivery for the result you choose. Choose the result that pays you, make sure Meta can see it, and give it enough data to learn.